Estimating neural response functions from fMRI.

Sukhbinder Kumar,William Penny

doi:10.3389/fninf.2014.00048

Abstract

This paper proposes a methodology for estimating Neural Response Functions (NRFs) from fMRI data. These NRFs describe non-linear relationships between experimental stimuli and neuronal population responses. The method is based on a two-stage model comprising an NRF and a Hemodynamic Response Function (HRF) that are simultaneously fitted to fMRI data using a Bayesian optimization algorithm. This algorithm also produces a model evidence score, providing a formal model comparison method for evaluating alternative NRFs. The HRF is characterized using previously established “Balloon” and BOLD signal models. We illustrate the method with two example applications based on fMRI studies of the auditory system. In the first, we estimate the time constants of repetition suppression and facilitation, and in the second we estimate the parameters of population receptive fields in a tonotopic mapping study.

Highlights

Functional Magnetic Resonance Imaging is a well established technique for the non-invasive mapping of human brain function (Frackowiak et al, 2003)
An Neural Response Functions (NRFs) specifies how neuronal activity is related to stimulus characteristics and an Hemodynamic Response Function (HRF) specifies how Functional Magnetic Resonance Imaging (fMRI) data is related to neuronal activity
We report results on an exponential “item-lag” model, in which neuronal responses were modeled using Equation (2), k indexes the four stimulus types (HC, click train (CT), Regular Interval noise stimuli (Noise) (RIN), Noise), and rk encodes the number of item repeats since the first stimulus of that type in the epoch

Summary

Introduction

Functional Magnetic Resonance Imaging (fMRI) is a well established technique for the non-invasive mapping of human brain function (Frackowiak et al, 2003). Analysis of fMRI data most often proceeds by modeling the neuronal correlates of single events as delta functions or boxcars These form event streams which are convolved with assumed Hemodynamic Response Functions (HRFs) to create regressors for General Linear Models (GLMs). This forms the basis of the widely used Statistical Parametric Mapping (SPM) approach (Friston et al, 2007). This comprises two mappings (1) a Neural Response Function (NRF) which maps stimulus characteristics to neural responses and (2) an HRF which maps neural responses to fMRI data. This algorithm has the added benefit of producing a model evidence score which we will use to provide a formal model comparison method (Penny, 2012) for evaluating alternative NRFs

Objectives

Methods

Results

Conclusion